Pneumonia is an enormous public health problem in the United States (US), accounting for 63,000 deaths and 1.2 million hospitalizations annually. A rapid and accurate assessment of pneumonia severity at the time of initial diagnosis is essential to guide optimal management. However, accurately identifying the severity of pneumonia based on clinical features alone is challenging and both overestimation and underestimation of severity are common, resulting in suboptimal patient outcomes and unnecessary healthcare expenditures. Salivary cortisol, a biomarker of physiological stress, has recently been proposed as a marker of illness severity in acute infections. By indicating the intensity of the body's adrenocortical response, salivary cortisol measurement may be a useful technique to assess the systemic stress associated with acute pneumonia and the patient's vulnerability to severe outcomes. However, significant further research is needed to understand its predictive performance for clinically-important outcomes. In this proposal, we will robustly evaluate salivary cortisol as a predictor of clinical outcomes in adults with community-acquired pneumonia (CAP) and develop and internally validate a clinical prediction model including salivary cortisol for CAP severity, as measured by death, septic shock, respiratory failure, intensive care unit admission, and prolonged hospital length of stay (Aim 1). Next, we will determine if the new salivary cortisol model improves severity assessment beyond other tools currently available (Aim 2). For Aims 1 and 2, we will utilize banked samples and extensive patient data we previously collected in the Centers for Disease Control and Prevention (CDC) Etiology of Pneumonia in the Community (EPIC) Study, a recent multisite prospective cohort study. Next, we will perform external validation of the salivary cortisol model in a separate cohort of CAP patients (Aim 3). Finally, in a biomarker discovery analysis using a metabolomics approach, we will identify other salivary metabolites for biomarker development to complement salivary cortisol (Aim 4). To accomplish Aims 3 and 4, we will recruit a new prospective cohort of CAP patients. The overarching objective of this K23 Award is to facilitate my development into an innovative, federally-funded, patient-oriented researcher. I will be closely mentored by seven distinguished researchers, led by primary mentor Kathryn M. Edwards, a world-renowned infectious diseases scientist who has successfully mentored dozens of junior investigators to thriving academic careers. Additionally, I will complete advanced coursework carefully selected to provide a deep understanding of the scientific methods most important for my career, including epidemiological approaches for designing prospective clinical studies, statistical methods for evaluating clinical prediction models and biomarkers, and translational metabolomics research. As a faculty member at Vanderbilt University, I am immersed in a highly collaborative academic environment with devoted mentors, engaging colleagues, a diverse patient population, elite educational programs, and a profound institutional commitment to foster my development.